FEA of hyperelastic arteries in ANSYS APDL

[Killian Chellar]

Hi all,

I am attempting to model an artery under internal pressure in ANSYS APDL.. This is my current attempt..

1. select solid 8 node 183 element type (and specify thickness)
2. select 5 parameter Mooney-Rivlin (for which I have constants)
3. create quarter area model dimensions
4. map mesh with quad area elements
5. Apply symmetry Boundary Conditions to artery inner cut lines
6. Apply pressure load on internal area line
7. Activate large deformation (NLGEOM,ON)
8. Solve using static analysis

This approach works fine for linear-elastic materials, but I cannot understand why it yields no results in this situation?

ANY ADVICE??
Thanks in advance

 

[Chestermiller]

Hi Killian Chellar. Welcome to Physics Forums!

I don't have any direct experience with FEA, but I have lots of experience with numerical analysis. I have some thoughts and some questions.

My understanding is that you are solving a large set of non-linear algebraic equations. You experienced no problems when you considered cases of linear elastic behavior, but could not get any meaningful results when you switched to Mooney Rivlin.

Questions:
1. Was the large deformation description activated when you solved the linearly elastic case.
2. Do you apply the load all in one shot, or do you solve the problem completely several times in succession using small increments in the applied load, starting from the solution for the previous load?
3. Have you tried cutting back on the Mooney-Rivlin parameters so that the model is nearly linear, solving that problem, then increasing the Mooney-Rivlin parameters, starting with the solution for the previous set of parameters.
4. Have you tried doing the above two tricks with both the load and the material parameters at the same time.

What I'm saying is that, with the Mooney-Rivlin parameters, maybe the equations may be too non-linear, and you haven't been using a good enough initial guess to converge to a solution. One way of getting around this is to sneak up on the load, and another way is to sneak up on the material properties, and another way is to sneak up on both.

Chet

 

[Chestermiller]

Incidentally, this sounds like a 1D problem, with only one independent variable, the radial location. So the equations should be expressible as ODEs, and can be solved by ODE methods. Is it correct that you are just dealing with a tube made out of material that behaves according to the Mooney-Rivlin equation, and you are just applying internal pressure to the tube? If so, then such a problem is much easier to solve than by using finite element.

Chet